Guideline: The same data should produce the same visualization (D->V).
Source: G. Kindlmann and C. Scheidegger. “An Algebraic Process for Visualization Design”. http://vis.cs.ucdavis.edu/vis2014papers/TVCG/papers/2181_20tvcg12-kindlmann-2346325.pdf
I agree with this guideline in an illusive manner. But I am not sure what “same visualization” really means. If one replaces a bar chart with a pie chart, one high-res display with a low-res one, one view of a 3D scene with another view, one iso-value with another, and so on, do they lead to a different or the same visualization? If the function D->V is absolutely fixed, different people may see different visualizations due to the difference of their visual systems or environments (e.g., color blindness, and lighting in the room).